Parallel on-chip gene synthesis and application to optimization of protein expression

Low-cost, high-throughput gene synthesis and precise control of protein expression are of critical importance to synthetic biology and biotechnology. Here we describe the development of an on-chip gene synthesis technology, which integrates on a single microchip the synthesis of DNA oligonucleotides using inkjet printing, isothermal oligonucleotide amplification and parallel gene assembly. Use of a mismatch-specific endonuclease for error correction results in an error rate of ~0.19 errors per kb. We applied this approach to synthesize pools of thousands of codon-usage variants of lacZα and 74 challenging Drosophila protein antigens, which were then screened for expression in Escherichia coli. In one round of synthesis and screening, we obtained DNA sequences that were expressed at a wide range of levels, from zero to almost 60% of the total cell protein mass. This technology may facilitate systematic investigation of the molecular mechanisms of protein translation and the design, construction and evolution of macromolecular machines, metabolic networks and synthetic cells.     

Rational de novo gene synthesis by rapid polymerase chain assembly (PCA) and expression of endothelial protein-C and thrombin receptor genes

The assembly of synthetic oligonucleotides into genes and genomes is an important methodology. Several methodologies for such synthesis have been developed, but they have two drawbacks: (1) the processes are slow and (2) the error frequencies are high (typically 1-3 errors/kb of DNA). Thermal damage is a major contributor to biosynthetic errors. In this paper, we elucidate the advantages of rapid gene synthesis by polymerase chain assembly (PCA) when used in combination with smart error control strategies. We used a high-speed thermocycler (PCRJet) to effectively minimize thermal damage and to perform rapid assembly of synthetic oligonucleotides to construct two different genes: endothelial protein C receptor (EPCR) and endothelial cell thrombin receptor, thrombomodulin (TM). First, the intact EPCR gene (EPCR-1, 612 bp) and a mutant EPCR-2 (576 bp) that lacked 4 N-linked glycosylation sites were constructed from 35 and 33 oligonucleotides, respectively. Next, for direct error comparison, another longer gene, the 1548 bp TM gene was constructed from 87 oligonucleotides by both rapid and conventional PCA. The fidelity and accuracy of the synthetic genes generated in this manner were confirmed by sequencing. The combined steps of PCA and DNA amplification are completed in about 10 and 22 min for EPCR-1, 2 and TM genes, respectively with comparable low errors in the DNA sequence. Furthermore, we subcloned synthetic TM, EPCR-1, EPCR-2 and native EPCR-1 (amplified from cDNA) into a Pichia pastoris expression vector to evaluate the expression ability, and to compare them with the native gene. Here, we illustrate that the synthetic genes, assembled by rapid PCA, successfully directed the expression of functional proteins. And, importantly, the synthetic and the native genes expressed proteins with the same efficiency.     

Removal of mismatched bases from synthetic genes by enzymatic mismatch cleavage

The success of long polynucleotide de novo synthesis is largely dependent on the quality and purity of the oligonucleotides used. Generally, the primary product of any synthesis reaction is directly cloned, and clones with correct products have to be identified. In this study, a novel strategy has been established for removing undesired sequence variants from primary gene synthesis products. Single base-pair mismatches, insertions and deletions were cleaved with specific endonucleases. Three different enzymes—T7 endonuclease I, T4 endonuclease VII and Escherichia coli endonuclease V—have been tested. As a model, a synthetic polynucleotide encoding the bacterial chloramphenicol-acetyltransferase (cat) was synthesized using different methods for one step polynucleotide synthesis based on ligation of oligonucleotides. The influence of enzymatic mismatch cleavage (EMC) as an error correction step on the frequency of correct products was analyzed by functional cloning of the synthetic cat and comparing the error rate with that of untreated products. Significant reduction of all mutation types was observed. Statistical analysis revealed that the T4 and E.coli endonucleases reduced the occurrence of mutations in cloned synthetic gene products. The EMC treatment was successful especially in the removal of deletions and insertions from the primary ligation products.     

Characterization of Arabidopsis thaliana mismatch specific endonucleases: application to mutation discovery by TILLING in pea

Scanning DNA sequences for mutations and polymorphisms has become one of the most challenging, often expensive and time-consuming obstacles in many molecular genetic applications, including reverse genetic and clinical diagnostic applications. Enzymatic mutation detection methods are based on the cleavage of heteroduplex DNA at the mismatch sites. These methods are often limited by the availability of a mismatch-specific endonuclease, their sensitivity in detecting one allele in a pool of DNA and their costs. Here, we present detailed biochemical analysis of five Arabidopsis putative mismatch-specific endonucleases. One of them, ENDO1, is presented as the first endonuclease that recognizes and cleaves all types of mismatches with high efficiency. We report on a very simple protocol for the expression and purification of ENDO1. The ENDO1 system could be exploited in a wide range of mutation diagnostic tools. In particular, we report the use of ENDO1 for discovery of point mutations in the gibberellin 3beta-hydrolase gene of Pisum sativum. Twenty-one independent mutants were isolated, five of these were characterized and two new mutations affecting internodes length were identified. To further evaluate the quality of the mutant population we screened for mutations in four other genes and identified 5-21 new alleles per target. Based on the frequency of the obtained alleles we concluded that the pea population described here would be suitable for use in a large reverse-genetics project.     

Specific A/G-to-C.G mismatch repair in Salmonella typhimurium LT2 requires the mutB gene product.

An assay has been developed that permits analysis of repair of A/G mismatches to C.G base pairs in cell extracts of Salmonella typhimurium LT2. This A/G mismatch repair is independent of ATP, dam methylation, and mutS gene function. The gene product of mutB has been shown to be involved in the dam-independent pathway through the in vitro assay. Moreover, specific DNA-protein complexes and an endonuclease can be detected in S. typhimurium extracts by using DNA fragments containing an A/G mismatch. These activities are not observed with substrates which have a T/G mismatch or no mismatch. The S. typhimurium endonuclease, like the A/G endonuclease found in Escherichia coli (A-L. Lu and D.-Y. Chang, Cell 54:805-812, 1988), makes incisions at the first phosphodiester bond 3' to and the the second phosphodiester bond 5' to the dA of the A/G mismatch. No incision site was detected on the other DNA strand. Extracts prepared from mutB mutants cannot form A/G mismatch-specific DNA-protein complexes and do not contain the A/G endonuclease activity. Thus the A/G mismatch specific binding and nicking activities are probably involved in the A/G mismatch repair pathway. Preliminary analysis of the mutational spectrum of the mutB strain has indicated that this mutator allele causes an increase in C.G-to-A.T transversions without affecting the frequencies of other transversion or transition events. In addition, the mutB gene has been mapped to the 64-min region of the S. typhimurium chromosome. Together, this biochemical and genetic evidence suggests that the mutB gene product of S. typhimurium is the homolog of the E. coli micA (and/or mutY) gene product.     

Mutation detection using ENDO1: Application to disease diagnostics in humans and TILLING and Eco-TILLING in plants

Background  

Most enzymatic mutation detection methods are based on the cleavage of heteroduplex DNA by a mismatch-specific endonuclease at mismatch sites and the analysis of the digestion product on a DNA sequencer. Important limitations of these methods are the availability of a mismatch-specific endonuclease, their sensitivity in detecting one allele in pool of DNA, the cost of the analysis and the ease by which the technique could be implemented in a standard molecular biology laboratory.     

Inteins of Thermococcus fumicolans DNA polymerase are endonucleases with distinct enzymatic behaviors

The DNA polymerase gene of Thermococcus fumicolans harbors two intein genes. Both inteins have been produced in Escherichia coli and purified either as naturally spliced products from the expression of the complete DNA polymerase gene or directly from the cloned inteins genes. Both recombinant inteins exhibit endonuclease activity, with an optimal temperature of 70 degrees C. The Tfu pol-1 intein, which belongs to the Psp KOD pol-1 allelic family, recognizes and cleaves a minimal sequence of 16 base pairs (bp) on supercoiled DNA with either Mn(2+) or Mg(2+) as cofactor. It cleaves linear DNA only with Mn(2+) and requires a 19-bp minimal recognition sequence. The Tfu pol-2 intein, which belongs to the Tli pol-2 allelic family, is a highly active homing endonuclease using Mg(2+) as cofactor. Its minimal recognition and cleavage site is 21 bp long either on linear or circular DNA substrates. Its endonuclease activity is strongly inhibited by the 3' digestion product, which remains bound to the enzyme after the cleavage reaction. According to current nomenclature, these endonucleases were named PI-TfuI and PI-TfuII. These two inteins thus exhibit different requirements for metal cofactor and substrate topology as well as different mechanism of action.     

EST clustering error evaluation and correction

MOTIVATION: The gene expression intensity information conveyed by (EST) Expressed Sequence Tag data can be used to infer important cDNA library properties, such as gene number and expression patterns. However, EST clustering errors, which often lead to greatly inflated estimates of obtained unique genes, have become a major obstacle in the analyses. The EST clustering error structure, the relationship between clustering error and clustering criteria, and possible error correction methods need to be systematically investigated. RESULTS: We identify and quantify two types of EST clustering error, namely, Type I and II in EST clustering using CAP3 assembling program. A Type I error occurs when ESTs from the same gene do not form a cluster whereas a Type II error occurs when ESTs from distinct genes are falsely clustered together. While the Type II error rate is <1.5% for both 5' and 3' EST clustering, the Type I error in the 5' EST case is approximately 10 times higher than the 3' EST case (30% versus 3%). An over-stringent identity rule, e.g., P >/= 95%, may even inflate the Type I error in both cases. We demonstrate that approximately 80% of the Type I error is due to insufficient overlap among sibling ESTs (ISO error) in 5' EST clustering. A novel statistical approach is proposed to correct ISO error to provide more accurate estimates of the true gene cluster profile.     

The use of thioglycolate to distinguish between 3'' AP (apurinic/apyrimidinic) endonucleases and AP lyases.

Addition of thioglycolate and DEAE-Sephadex chromatography were used to analyze the cleavage of the C(3')-O-P bond 3' to AP (apurinic/apyrimidinic) sites in DNA and to distinguish between a mechanism of hydrolysis (which would allow the nicking enzyme to be called 3' AP endonuclease) or beta-elimination (so that the nicking enzyme should be called AP lyase). For this purpose, DNA labelled in the AP sites was first cleaved by rat-liver AP endonuclease, then with the 3' nicking catalyst in the presence of thioglycolate and the reaction products were analyzed on DEAE-Sephadex: deoxyribose-5-phosphate (indicating a 3' cleavage by hydrolysis) and the thioglycolate:unsaturated sugar-5-phosphate adduct (indicating a cleavage by beta-elimination) are well separated allowing to eventually easily discard the hypothesis of a hydrolytic process and the appellation of 3' AP endonuclease. We have shown that addition of thioglycolate to the unsaturated sugar resulting from nicking the C(3')-O-P bond 3' to AP sites by beta-elimination is an irreversible reaction. We have also shown that the thioglycolate must be present from the beginning of the reaction with the nicking catalyst to prevent the primary 5' product of the beta-elimination reaction from undergoing other modifications that complicate the interpretation of the results.     

Mutation detection using Surveyor nuclease

We have developed a simple and flexible mutation detection technology for the discovery and mapping of both known and unknown mutations. This technology is based on a new mismatch-specific DNA endonuclease from celery, Surveyor nuclease, which is a member of the CEL nuclease family of plant DNA endonucleases. Surveyor nuclease cleaves with high specificity at the 3' side of any mismatch site in both DNA strands, including all base substitutions and insertion/deletions up to at least 12 nucleotides. Surveyor nuclease technology involves four steps: (i) PCR to amplify target DNA from both mutant and wild-type reference DNA; (ii) hybridization to form heteroduplexes between mutant and wild-type reference DNA; (iii) treatment of annealed DNA with Surveyor nuclease to cleave heteroduplexes; and (iv) analysis of digested DNA products using the detection/separation platform of choice. The technology is highly sensitive, detecting rare mutants present at as low as 1 in 32 copies. Unlabeled Surveyor nuclease digestion products can be analyzed using conventional gel electrophoresis or high-performance liquid chromatography (HPLC), while end labeled digestion products are suitable for analysis by automated gel or capillary electrophoresis. The entire protocol can be performed in less than a day and is suitable for automated and high-throughput procedures.     

The site-specific cleavage of synthetic Holliday junction analogs and related branched DNA structures by bacteriophage T7 endonuclease I

Various branched DNA structures were created from synthetic, partly complementary oligonucleotides combined under annealing conditions. Appropriate mixtures of oligonucleotides generated three specific branched duplex DNA molecules: (i) a Holliday junction analog having a fixed (immobile) crossover bounded by four duplex DNA branches, (ii) a similar Holliday junction analog which is capable of limited branch migration and, (iii) a Y-junction, with three duplex branches and fixed branch point. Each of these novel structures was specifically cleaved by bacteriophage T7 gene 3 product, endonuclease I. The cleavage reaction "resolved" the two Holliday structure analogs into pairs of duplex DNA products half the size of the original molecules. The point of cleavage in the fixed-junction molecules was predominantly one nucleotide removed to the 5' side of the expected crossover position. Multiple cleavage positions were mapped on the Holliday junction with the mobile, or variable, branch point, to sites consistent with the unrestricted movement of the phosphodiester crossover within the region of limited dyad symmetry which characterizes this molecule. Based on the cleavage pattern observed with this latter substrate, the enzyme displayed a modest degree of sequence specificity, preferring a pyrimidine on the 3' side of the cleavage site. Branched molecules that were partial duplexes (lower order complexes which possessed single-stranded as well as duplex DNA branches) were also substrates for the enzyme. In these molecules, the cleaved phosphodiester bonds were in duplex regions only and predominantly one nucleotide to the 5' side of the branch point. The phosphodiester positions 5' of the branch point in single-stranded arms were not cleaved. Under identical reaction conditions, individually treated oligonucleotides were completely refractory. Thus, cleavage by T7 endonuclease I displays great structural specificity with an efficiency that can vary slightly according to the DNA sequence.     

A Mutant Brassica napus (Canola) Population for the Identification of New Genetic Diversity via TILLING and Next Generation Sequencing

We have generated a Brassica napus (canola) population of 3,158 EMS-mutagenised lines and used TILLING to demonstrate that the population has a high enough mutation density that it will be useful for identification of mutations in genes of interest in this important crop species. TILLING is a reverse genetics technique that has been successfully used in many plant and animal species. Classical TILLING involves the generation of a mutagenised population, followed by screening of DNA samples using a mismatch-specific endonuclease that cleaves only those PCR products that carry a mutation. Polyacrylamide gel detection is then used to visualise the mutations in any gene of interest. We have used this TILLING technique to identify 432 unique mutations in 26 different genes in B. napus (canola cv. DH12075). This reflects a mutation density ranging from 1/56 kb to 1/308 kb (depending on the locus) with an average of 1/109 kb. We have also successfully verified the utility of next generation sequencing technology as a powerful approach for the identification of rare mutations in a population of plants, even in polyploid species such as B. napus. Most of the mutants we have identified are publically available.     

Screening Human Genes for Small Alterations Performing an Enzymatic Cleavage Mismatched Analysis (ECMA) Protocol

Many human diseases are caused by small alterations in the genes and in the majority of cases sophisticated protocols are required for their detection. In this study we estimated the efficacy of an enzymatic protocol, which using a new mismatch-specific DNA plant endonuclease from celery (CEL family) recognizes and cleaves mismatched alleles between mutant and normal PCR products. The protocol was standardized on a variety of known mutations, in 11 patients with cystic fibrosis (CF), Fabry’s disease (FD), steroid 21-hydroxylase deficiency (21-HD) and Duchenne/Becker muscular dystrophy (DMD/BMD). The method does not require special equipment, labeling or standardization for every PCR product, since conditions of heteroduplex formation and enzyme digestion are universal for all products. The results showed that the method is rapid, effective, safe, reliable, and very simple, as the mutations are visualized on agarose or nusieve/agarose gels. The protocol was furthermore evaluated in three DMD patients with the detection of three alterations which after sequencing, were characterized as disease causative mutations. The proposed assay, which was applied for the first time in a variety of monogenic disorders, indicates that point mutation identification is feasible in any conventional molecular lab even for cases, where other techniques have failed.     

Splicing of a yeast proline tRNA containing a novel suppressor mutation in the anticodon stem

The intron-containing proline tRNAUGG genes in Saccharomyces cerevisiae can mutate to suppress +1 frameshift mutations in proline codons via a G to U base substitution mutation at position 39. The mutation alters the 3' splice junction and disrupts the bottom base-pair of the anticodon stem which presumably allows the tRNA to read a four-base codon. In order to understand the mechanism of suppression and to study the splicing of suppressor pre-tRNA, we determined the sequences of the mature wild-type and mutant suppressor gene products in vivo and analyzed splicing of the corresponding pre-tRNAs in vitro. We show that a novel tRNA isolated from suppressor strains is the product of frameshift suppressor genes. Sequence analysis indicated that suppressor pre-tRNA is spliced at the same sites as wild-type pre-tRNA. The tRNA therefore contains a four-base anticodon stem and nine-base anticodon loop. Analysis of suppressor pre-tRNA in vitro revealed that endonuclease cleavage at the 3' splice junction occurred with reduced efficiency compared to wild-type. In addition, reduced accumulation of mature suppressor tRNA was observed in a combined cleavage and ligation reaction. These results suggest that cleavage at the 3' splice junction is inefficient but not abolished. The novel tRNA from suppressor strains was shown to be the functional agent of suppression by deleting the intron from a suppressor gene. The tRNA produced in vivo from this gene is identical to that of the product of an intron+ gene, indicating that the intron is not required for proper base modification. The product of the intron- gene is a more efficient suppressor than the product of an intron+ gene. One interpretation of this result is that inefficient splicing in vivo may be limiting the steady-state level of mature suppressor tRNA.     

Intron 1 sequences are required for pancreatic expression of the human proglucagon gene

The mammalian proglucagon gene is expressed in pancreatic islet A-cells, intestinal L-cells, and select neurons of the brain, where posttranslational processing results in the liberation of a unique profile of peptides. Despite the importance of proglucagon-derived peptides in human biology, little is known about the regulation of the human gene, as the rat gene has been the preferred model for understanding the regulation of proglucagon gene expression. Previously, we have shown that although the immediate promoter region of the rat proglucagon gene is sufficient for expression in pancreatic islet cells, the homologous human proglucagon promoter sequences are not sufficient. We have now used a comparative genomic approach to identify noncoding sequences near the human proglucagon gene that are conserved among mammals, and thus potentially are regulatory sequences. Our alignments identified three evolutionarily conserved noncoding regions (ECR), one is the immediate promoter region (ECR1), the second is about 5 kb 5' to the mRNA start site (ECR2), and the third is near the 3' end of the first intron (ECR3). Our in vitro transient transfection assays with reporter gene constructs that include the human ECR3 support expression in rodent islet cell lines. Complementary studies with transgenic mice possessing a reporter gene regulated by a human proglucagon gene promoter-intron 1 (including ECR3) sequences express the reporter gene in the pancreas, as well as the intestine and selected neurons. These studies suggest that conserved sequences within intron 1 of the human proglucagon gene are important for expression in the pancreas.     

Protein-mediated error correction for de novo DNA synthesis

The availability of inexpensive, on demand synthetic DNA has enabled numerous powerful applications in biotechnology, in turn driving considerable present interest in the de novo synthesis of increasingly longer DNA constructs. The synthesis of DNA from oligonucleotides into products even as large as small viral genomes has been accomplished. Despite such achievements, the costs and time required to generate such long constructs has, to date, precluded gene-length (and longer) DNA synthesis from being an everyday research tool in the same manner as PCR and DNA sequencing. A critical barrier to low-cost, high-throughput de novo DNA synthesis is the frequency at which errors pervade the final product. Here, we employ a DNA mismatch-binding protein, MutS (from Thermus aquaticus) to remove failure products from synthetic genes. This method reduced errors by >15-fold relative to conventional gene synthesis techniques, yielding DNA with one error per 10000 base pairs. The approach is general, scalable and can be iterated multiple times for greater fidelity. Reductions in both costs and time required are demonstrated for the synthesis of a 2.5 kb gene.     

Escherichia coli endonuclease III is not an endonuclease but a beta-elimination catalyst.

The oligonucleotide [5'-32P]pdT8d(-)dTn, containing an apurinic/apyrimidinic (AP) site [d(-)], yields three radioactive products when incubated at alkaline pH: two of them, forming a doublet approximately at the level of pdT8dA when analysed by polyacrylamide-gel electrophoresis, are the result of the beta-elimination reaction, whereas the third is pdT8p resulting from beta delta-elimination. The incubation of [5'-32P]pdT8d(-)dTn, hybridized with poly(dA), with E. coli endonuclease III yields two radioactive products which have the same electrophoretic behaviour as the doublet obtained by alkaline beta-elimination. The oligonucleotide pdT8d(-) is degraded by the 3'-5' exonuclease activity of T4 DNA polymerase as well as pdT8dA, showing that a base-free deoxyribose at the 3' end is not an obstacle for this activity. The radioactive products from [5'-32P]pdT8d(-)dTn cleaved by alkaline beta-elimination or by E. coli endonuclease III are not degraded by the 3'-5' exonuclease activity of T4 DNA polymerase. When DNA containing AP sites labelled with 32P 5' to the base-free deoxyribose labelled with 3H in the 1' and 2' positions is degraded by E. coli endonuclease VI (exonuclease III) and snake venom phosphodiesterase, the two radionuclides are found exclusively in deoxyribose 5-phosphate and the 3H/32P ratio in this sugar phosphate is the same as in the substrate DNA. When DNA containing these doubly-labelled AP sites is degraded by alkaline treatment or with Lys-Trp-Lys, followed by E. coli endonuclease VI (exonuclease III), some 3H is found in a volatile compound (probably 3H2O) whereas the 3H/32P ratio is decreased in the resulting sugar phosphate which has a chromatographic behaviour different from that of deoxyribose 5-phosphate. Treatment of the DNA containing doubly-labelled AP sites with E. coli endonuclease III, then with E. coli endonuclease VI (exonuclease III), also results in the loss of 3H and the formation of a sugar phosphate with a lower 3H/32P ratio that behaves chromatographically as the beta-elimination product digested with E. coli endonuclease VI (exonuclease III). From these data, we conclude that E. coli endonuclease III cleaves the phosphodiester bond 3' to the AP site, but that the cleavage is not a hydrolysis leaving a base-free deoxyribose at the 3' end as it has been so far assumed. The cleavage might be the result of a beta-elimination analogous to the one produced by an alkaline pH or Lys-Trp-Lys. Thus it would seem that E. coli 'endonuclease III' is, after all, not an endonuclease.